Guest blog post by Dr Eyo Eyo
AI has directly impacted people’s lives and activities. This has subsequently resulted in significant concerns and questions about ethical considerations, trust, and legality.
It is worth emphasising that AI has brought numerous benefits and opportunities to businesses and organisations. Nonetheless, this has also entailed a greater deal of responsibility especially on how AI is used.
No doubt, perceptions are still quite mixed as to the degree of impact (positive or negative) that AI-enabled technologies can have on society, especially in the longer term. Recent data from research conducted by Accenture indicates that about 77 % of tech consumers believe that organisations must be made to account for the misuse of AI. In the same study, only about 35% of global tech consumers seem to trust how AI is deployed by businesses and organisations.
The influence of AI on social life and work has also been documented as studies have revealed how AI-driven decision-making can lead to biased outcomes, racial profiling in predictive algorithms used in policing and in sexist hiring decisions.
On the other hand, many other companies and organisations cannot even seem to see or recognise that there are problems or potential dangers in the first place. This is following a recent report from the Massachusetts Institute of Technology (MIT) Sloan Management Review and another well-known Consulting Group.
Notwithstanding, it should be a matter of priority that as AI technologies are being scaled up to capture business benefits, new regulations are made, and steps are taken to ensure adequate compliance. This is where responsible AI becomes very important.
It is no surprise that in recent years governments worldwide have begun to tighten regulations and rules targeting AI by encouraging and leading the increased adoption of ethical, responsible, and trustworthy AI initiatives.
Responsible AI in the educational context: Current realities
Responsible AI is concerned with the design, deployment, and execution of transparent, ethical, and accountable AI technology to reduce biases, promote fairness, equality, and aid in facilitating explainability and interpretability of outcomes, that are particularly pertinent in an educational context.
The arrival and proliferation of AI tools such as large language models (ChatGPT and similar transformers) and deep learning models(capable of generating images from text) have provoked some serious concerns and debate among educators worldwide, on the actual and potential consequences for teaching, learning and student assessment. Moreover, keeping pace with the potential dangers of each new AI development and breakthrough has posed a relentless challenge especially, in university education.
An important dilemma for tertiary institutions remains that industries are increasingly requiring graduates with knowledge and expertise in machine learning applications with a keen understanding of ethical and societal considerations combined with the ability to communicate.
Ethical awareness and shortcomings in education
Although still very limited in scope, the endeavours to establish guidelines, principles, and strategies for responsible AI that conform to essential human values and legal norms are promising foundations for responsible AI adoption and implementation in education.
At present, ethical considerations regarding AI and education only seem to primarily revolve around privacy, security, and ensuring the proper utilisation of personal data. Also, in environments where virtual assistants are becoming more prevalent in aiding students’ learning and teachers’ performance, there are increasing concerns about the impact of these assistants on learners.
The use of AI is linked to several shortcomings, and this includes but is not limited to the absence of references for information sources, biases in data and algorithms, concerns around intellectual property and copyright, and matters relating to privacy, data security, and fairness.
Mitigating potential harm and ensuring responsible AI in education
It is crucial to acknowledge that the application of AI for academic purposes could yield numerous potential benefits, such as enhanced efficiency in studies, personalised or customised learning, and novel methods of working. Hence, prohibiting or banning the use of new technologies, including AI tools, would prove ineffective.
Instead, the higher education sector should ensure to adapt its learning, teaching, and assessment methods by integrating AI more effectively and suitably. Universities must explore innovative means how to employ AI tools in a way that aligns with their mission, objectives, and principles, while also considering their legal framework and the wider implications for society, culture, and the economy.
The disruptions brought about by AI tools highlight the necessity to re-evaluate and reform teaching and assessment methodologies and practices. This does reinforce the emphasis across the sector which was acknowledged during the Covid-19 pandemic when coursework and genuine formative evaluation were deemed necessary.
It is important for university staff and students to engage in continuous formal discussions regarding the ethical, responsible, and transparent use of emerging technologies, including AI tools. Also, besides the revision of policies at the institutional level, guidance on day-to-day practices is also necessary. This should consider academic integrity, including the responsibility to recognise and reference the use of AI in academic and student work and the restriction of its use for certain types of learning and assessment.
Finally, universities must continue to ensure that the consequences of new technologies on our societies are examined and assessed and that graduates are adequately prepared for the evolving labour markets due to digital and new technologies, most especially artificial intelligence.